Range Characterization and Diet of the Chisana Caribou Herd, Alaska
Authors, etc.
Wrangell-St. Elias National Park and Preserve
Abstract
[to be written]
Introduction
[copied from LM’s preliminary report]
The Chisana Caribou Herd (CCH) is a small herd of woodland caribou located in northeastern Wrangell St. Elias (WRST) National Park and Preserve in Alaska and the Tetlin National Wildlife Refuge in the Yukon. During the early 1990s, many caribou herds in interior Alaska experienced population declines of varying severity (Valkenburg et al., 1996). From 1989 to 2001, the CCH population dropped from an estimated 1800 to 400 individuals (R. Farnell, unpublished data). The CCH became listed as a “Specially Protected” population in 2002 under the Yukon’s Federal Species at Risk Act. After the implementation of a captive rearing program and the development of an international, multiagency management plan, the CCH population is now stable and open to subsistence hunting. However, the current population is significantly below pre-decline levels. Several factors most likely contributed to the CCH population decline including predation, low recruitment, and poor habitat.
Lichens may be the greatest feature limiting CCH habitat quality. Lichens comprise up to 60-86% of the winter diets of Yukon woodland caribou herds, while moss makes up only 1-8%, as estimated from the microhistological analysis of fecal pellets (Farnell and Gardner, 2002). Relative to other Alaskan and Yukon caribou, the CCH appear to have a winter diet that is unusually and low in lichen (31%) and high in moss (51%). Overflight landcover assessments and collections from a limited number of plots suggest that lichen cover with the range is sparse and dominated by nutrient-poor Stereocaulon species (Clarke and Waterreus, 2012), an indicator of poor winter habitat for reindeer (Swanson and Barker, 1992). More data on lichen and moss cover with the CCH range are needed to corroborate these results. In addition to cover, lichen biomass is an important measurement of caribou forage quality. If overgrazing occurs, lichen biomass may not be correlated with lichen cover, and caribou have shown preference for regions with low lichen cover, but high lichen biomass (Collins et al., 2010). Lichen biomass is commonly estimated from mean lichen mat thickness (Lieb, 1994; Arseneault et al., 1997; Moen et al. 2007; Collins et al., 2010). Therefore, measuring lichen thickness and cover in the CCH range may capture key details for assessing winter habitat.
Summer forage is also an important component of caribou habitat. Low recruitment in caribou populations in the Southern Alaska Peninsula is thought to be due to a combination of low winter lichen availability and poor summer range quality (Post and Klein, 1999). Caribou summer diet is typically composed of grasses, sedges, forbs, and willows (Boertje, 1984). Manipulation experiments within the CCH range suggest that climate change may affect caribou summer forage quality and abundance (Lenart et al., 2002). In this study, we aim to establish baseline monitoring of vegetation within the CCH range. Secondly, we estimate lichen cover and biomass across the CCH range to assess habitat quality relative to those of other caribou, particularly the neighboring Kluane, Aishihik, and Nelchina herds. In this report, we describe results from first season of data collection, present preliminary findings, and offer recommendations for future study years.
Methods
Study Area
We focus on the core range of the CCH within WRST National Park and Preserve boundaries. This area forms the basin between the Wrangell, St. Elias, and Nutzotin Mountains, containing the Chisana and White Rivers and the historic town of Chisana. Study area limits were defined to include the bulk of CCH radio collar signals recorded between 1988 and 2015 (Fig. 1; J. Putera, unpublished data, 2015). The resulting area is 2795 km2 (1079 miles2 ) and contains moderate to high elevations, from 1000 - 2500 m (3281–8202 ft).
We chose to proportionally allocate sampling sites to six major ecotypes within the study area (Table 1). Major ecotypes were developed by consolidating the detailed ecotypes mapped in WRST by Jorgenson et al. (2008) using satellite imagery. We grouped Jorgenson et al. ecotypes into groups based on the similarity of their definitions and geographic size (Table 1; Fig. 2). We included all ecotypes except for glaciated barrens. In our study region, the three most common ecotypes by percent cover are Alpine Dwarf Shrub/Barrens (28.1%), Subalpine Wood/Shrub (28.1%), and Alpine Sedge-Dwarf Shrub (24.3%). Riverine, Alpine Meadow, and Forest ecotypes make up the remaining 19.4% of the region.
Table 1. Ecotypes of Chisana caribou herd range, ranked by percent cover.
| Major ecotype | Included detailed ecotypes* | Distinguishing features | Landcover | |
| 1 | Alpine Dwarf-Shrub/Barrens | Alpine Dryas Dwarf Shrub Alpine Barrens Alpine Ericaeous Dwarf Shrub |
Dryas spp. and other evergreen dwarf shrubs dominant, trees or tall shrubs absent, and rare graminoids; occurs at high elevations | 28.1% |
| 2 | Subalpine Wood/Shrub | Subalpine Willow-Birch Shrub Subalpine Spruce Woodland |
Characterized by tall (> 0.2 m) deciduous shrubs; few sedges and grasses; tree cover < 20% | 28.1% |
| 3 | Alpine Sedge-Dwarf Shrub | Alpine Sedge-Dwarf Willow Meadow | Deciduous shrubs dominant, but with lower frequencies than in ecotype 2; sedges co-dominant; trees absent; relatively high species diversity | 24.3% |
| 4 | Riverine | Riverine Circumalkaline Barrens Riverine Sandy Willow Shrub Riverine Dryas Dwarf Shrub Riverine Tall Alder Shrub Riverine Low Silverberry Shrub |
On or near braided streams; deciduous or evergreen dwarf shrubs are dominant and trees uncommon | 7.5% |
| 5 | Alpine Meadow | Alpine Tussock Meadow Alpine Sedge Meadow |
Tussocks > 25% and sedge cover approaching 60%; dwarf shrubs can be co-dominant; no trees; wet soil | 7.3% |
| 6 | Forest | Subalpine Poplar Forest Riverine White Spruce Forest Riverine Gravelly Poplar Forest Riverine Spruce-Poplar Forest |
Common on or near overbank deposits; characterized by > 20% tree cover and thick mossy floor | 4.6% |
We generated 56 random points for site locations within the Chisana caribou herd range (Appendix A). We included a 3000 m buffer between sites to ensure an even geographic distribution throughout the study area. Points that fell within rivers were moved to the nearest accessible riverine ecotype and points on excessively steep terrain were eliminated. We re-assigned ecotypes after field data collection using the formal WSRT ecotype classification key (summarized in Table 1; Jorgenson et al., 2008). We visited 23 sites in July, 2015 and 10 sites in July, 2016. Sites were accessed by helicopter or foot using a Trimble or Garmin GPS for navigation (Fig. 3). Permanent plots were established at all sites except those falling with a proposed Archeological District around Wiki Peak. Permanent plots were marked with a 60 cm-long rebar located at plot centers. At each site, we described physical landscape characteristics such as slope position, drainage, and geomorphic disturbance.
Figure 1. Defining the core range of the Chisana caribou herd in northeastern WRST Park and Preserve. Points are radio collar signals recorded from 1988 to 2015 during summer or winter seasons. The core range (yellow border) was selected to include the majority of radio signals within WRST (green border).
Figure 2. Major ecotypes within the Chisana caribou range in northeastern WRST Park and Preserve (Table 1). Map based on ecotype landcover developed by Jorgenson et al. (2008).
RE-DO THIS MAP WITH 2016 SITES
Figure 3. Vegetation sampling sites in the Chisana caribou herd range. Sites visited in 2015 (green) and proposed for future study years (red) are labeled with site IDs. See Appendix A for the geographic locations of all sites.
Figure 6. Map of sampling plots.
Fig 4. Plot design schematic showing locations of transects.
Ecotypes
Table 1. Ecotypes of Chisana caribou herd range, ranked by percent cover.
| Major ecotype | Included detailed ecotypes* | Distinguishing features | Landcover | |
| 1 | Alpine Dwarf-Shrub/Barrens | Alpine Dryas Dwarf Shrub Alpine Barrens Alpine Ericaeous Dwarf Shrub |
Dryas spp. and other evergreen dwarf shrubs dominant, trees or tall shrubs absent, and rare graminoids; occurs at high elevations | 28.1% |
| 2 | Subalpine Wood/Shrub | Subalpine Willow-Birch Shrub Subalpine Spruce Woodland |
Characterized by tall (> 0.2 m) deciduous shrubs; few sedges and grasses; tree cover < 20% | 28.1% |
| 3 | Alpine Sedge-Dwarf Shrub | Alpine Sedge-Dwarf Willow Meadow | Deciduous shrubs dominant, but with lower frequencies than in ecotype 2; sedges co-dominant; trees absent; relatively high species diversity | 24.3% |
| 4 | Riverine | Riverine Circumalkaline Barrens Riverine Sandy Willow Shrub Riverine Dryas Dwarf Shrub Riverine Tall Alder Shrub Riverine Low Silverberry Shrub |
On or near braided streams; deciduous or evergreen dwarf shrubs are dominant and trees uncommon | 7.5% |
| 5 | Alpine Meadow | Alpine Tussock Meadow Alpine Sedge Meadow |
Tussocks > 25% and sedge cover approaching 60%; dwarf shrubs can be co-dominant; no trees; wet soil | 7.3% |
| 6 | Forest | Subalpine Poplar Forest Riverine White Spruce Forest Riverine Gravelly Poplar Forest Riverine Spruce-Poplar Forest |
Common on or near overbank deposits; characterized by > 20% tree cover and thick mossy floor | 4.6% |
Why only one plot in Forest? Merkhofer: “We found that Forest ecotypes made for difficult helicopter landing access and subsequently sampled only one site from this ecotype. Future sampling should be targeted to these under-sampled ecotypes.”
| Ecotype | n |
|---|---|
| Alpine Dwarf Shrub/Barrens | 11 |
| Subalpine Wood/Shrub | 9 |
| Alpine Sedge-Dwarf Shrub | 5 |
| Alpine Meadow | 4 |
| Riverine | 3 |
| Forest | 1 |
Vegetation Cover
We recorded vegetation along three 20 m-long transects oriented at 0, 120, and 240° from the plot center (Figure 4). Each transect was photographed and measured for slope. Transects were systematically sampled at 50 cm intervals using the line-point intercept method for 40 observations per transect. At each point, we recorded all intercepted overstory and understory taxa, soil surface cover, and lichen height. We identified taxa to the species or genus-level. We considered forage lichen as the genera Cladina, Cladonia (excluding species with cup-shaped podetia), Cetraria (excluding C. richardsonii), and Stereocaulon, following Collins et al. (2011). We estimated the percent cover of major life forms across the entire study area as the average of ecotype means weighted by ecotype area. Percent cover was calculated for each plot as total number of plant intercepts by species divided by 120 (3 transects X 40 sampling points).
[I didn’t do the below because I need help with it]
Following the methods recommended by Elzinga et al. (1998), we sequentially sampled percent cover values to test for a stable estimate of the overall population mean and standard deviation. We additionally used two-tailed significance tests to determine the necessary sample size for estimating percent cover for a single population and for detecting differences in percent cover between two time periods.
Lichen Thickness and Estimated Biomass
Lichen height was measured using a 3 mm-wide metal skewer inserted into the base of the lichen mat (Arseneault et al., 1997; Moen et al. 2007; Collins et al., 2010) at each of the 40 vegetation intercept points per transect. We calculated the mean lichen height of each plot as the average of all heights, including zero heights where no lichen was present. We estimated lichen biomass (kg/ha) for each plot from mean lichen height using the Moen et al. (2007) regression based on four forage lichen species:
\(B=1135.6T\),
where T is mean lichen mat thickness (cm).
Litter Depth and Composition
Litter depth was sampled from 10 cm-diameter soil plugs removed at the terminus of each transect. Our litter depth measurements include the height of any living mat-forming vegetation, such as moss.
Fecal Pellet Analysis
We collected fecal pellets from the range of the Chisana caribou herd in order to quantify what plants the Chisana caribou were feeding on. Collected fecal pellets were analyzed for taxa by XXX lab
Statistical Software
R (R Core Team, 2024) and R Studio software (Posit team, 2024) was used for all analyses.
Caribou Ecotype Preference
I performed a spatial join of the GPS collar data over the major ecotypes to see where caribou were spending their time. I used the ‘Join Attributes by Location’ tool in QGIS to add the ecotype to each GPS fix. I then summarized the number of pings by ecotype.
Results
Vegetation Cover
Foliar cover
*Unweighted by ecotype area.
Question: Merkhofer states “We estimated the percent cover of major life forms across the entire study area as the average of ecotype means weighted by ecotype area.” She did not give a formula. I’m uncertain how to weight means. The tables below give gover as the number of point intercept hits divided by the total number of available tries, unweighted by ecotype area.]
Results:
- Overall foliar plant cover in the vegetation plots we sampled was dominated by dwarf birch, sedges, Dryas and willow, species emblematic of alpine tundra.
| Taxa | % cover |
|---|---|
| Betula glandulosa | 17 |
| Unknown sedge | 12 |
| Dryas octopetala | 9 |
| Salix sp. | 8 |
| Unknown graminoid | 3 |
| Cassiope tetragona | 2 |
| Festuca altaica | 2 |
| Salix reticulata | 2 |
| Picea glauca | 2 |
| Empetrum nigrum | 2 |
| Ledum palustre | 2 |
| Salix arctica | 2 |
| Unknown grass | 2 |
| Unknown rush | 1 |
| Vaccinium uliginosum | 1 |
| Carex aquatilis | 1 |
| Carex bigelowii | 1 |
| Dryas integrifolia | 1 |
| Poa arctica | 1 |
| Equisetum scirpoides | 1 |
Basal Cover
*Unweighted by ecotype area.
| Taxa | % cover |
|---|---|
| Moss | 42 |
| Unknown sedge | 6 |
| Cladonia sp. | 5 |
| Stereocaulon sp. | 4 |
| Cetraria sp. | 3 |
| Empetrum nigrum | 3 |
| Peltigera sp. | 2 |
| Dryas octopetala | 1 |
| Unknown rush | 1 |
| Thamnolia sp. | 1 |
| Dryas integrifolia | 1 |
| Arctostaphylos uva-ursi | 1 |
| Unknown grass | 0 |
| Vaccinium vitis-idaea | 0 |
| Festuca altaica | 0 |
| Masonhalea richardsonii | 0 |
| Poa arctica | 0 |
| Salix sp. | 0 |
| Alectoria ochroleuca | 0 |
| Dactylina arctica | 0 |
| Betula glandulosa | 0 |
| Peltigera canina | 0 |
| Unknown graminoid | 0 |
| Picea glauca | 0 |
| Salix reticulata | 0 |
| Unknown lichen | 0 |
Lichen Layer Thickness and Biomass
NOTES on lichen biomass: I’m using LM’s conversion of lichen height to biomas “We estimated lichen biomass for each plot from mean lichen height using the Moen et al. (2007) regression based on four forage lichen species as B=1135.6T where T is mean lichen mat thickness (cm) and B is mean lichen mat biomass (kg/ha). ,1135.6 * AVG(IFNULL(v.LichenHeight,0)) as [Estimated lichen biomass (kg/ha)” without really knowing much about that methodology.
Also Moen et al. considered 4 leafy lichens. LM and JP did not distinguish any species of lichens in their lichen height measument at each point intercept (“We calculated the mean lichen height of each plot as the average of all heights, including zero heights where no lichen was present”).
Figure 7. Mean lichen thickness (cm) and estimated biomass (kg/Ha) by ecotype. Biomass estimated according to Moen et al.,(2007). Error bars represent one standard deviation. Note that only one plot was sampled in the Forest ecotype.
| Ecotype | Plots | Transects | Mean lichen thickness (cm) | SD | n |
|---|---|---|---|---|---|
| Alpine Dwarf Shrub/Barrens | 10 | 30 | 1.27 | 0.35 | 219 |
| Alpine Meadow | 4 | 12 | 1.52 | 0.36 | 164 |
| Alpine Sedge-Dwarf Shrub | 5 | 14 | 1.70 | 0.55 | 69 |
| Forest | 1 | 3 | 2.77 | 23 | |
| Riverine | 3 | 9 | 0.83 | 0.36 | 61 |
| Subalpine Wood/Shrub | 8 | 19 | 1.32 | 0.69 | 86 |
| Ecotype | Plots | Transects | Estimated biomass (kg/ha) | SD | n |
|---|---|---|---|---|---|
| Alpine Dwarf Shrub/Barrens | 10 | 30 | 1444.02 | 394.05 | 219 |
| Alpine Meadow | 4 | 12 | 1727.73 | 414.41 | 164 |
| Alpine Sedge-Dwarf Shrub | 5 | 14 | 1926.08 | 623.71 | 69 |
| Forest | 1 | 3 | 3150.06 | 23 | |
| Riverine | 3 | 9 | 945.99 | 403.44 | 61 |
| Subalpine Wood/Shrub | 8 | 19 | 1502.11 | 784.94 | 86 |
Figure 8. Mean lichen thickness (cm) by plot Error bars represent one standard deviation. Note that only one plot was sampled in the Forest ecotype.
Litter Depth and Composition
Each plot had 3 litter type and depth observations. This section assembles a list of all litter types encountered from the N,SE, and SW transects and counts how many observations there were for each to get at a percent of total litter observations per litter type.
| Litter type | Observations | % of transects |
|---|---|---|
| Moss | 43 | 43.43 |
| Soil | 12 | 12.12 |
| Herbaceous | 8 | 8.08 |
| Embedded litter | 5 | 5.05 |
| Dryas octopetala | 4 | 4.04 |
| Moss/lichen | 4 | 4.04 |
| Coniferous | 3 | 3.03 |
| Graminoid | 2 | 2.02 |
| Carex/forb | 1 | 1.01 |
| Lichen | 1 | 1.01 |
| Lichen/Moss | 1 | 1.01 |
| Moss/Sedge | 1 | 1.01 |
| Sedge | 1 | 1.01 |
| Shrub litter | 1 | 1.01 |
99 total transects.
Caribou Ecotype Preference
Ecotype preference
| Ecotype | GPS Fixes | Percent of GPS fixes |
|---|---|---|
| Subalpine Wood/Shrub | 4236 | 30 |
| Alpine Sedge-Dwarf Shrub | 4149 | 30 |
| Alpine Dwarf Shrub/Barrens | 2893 | 21 |
| Alpine Meadow | 2247 | 16 |
| Riverine | 257 | 2 |
| Forest | 147 | 1 |
Total GPS fixes = 13929
Ecotype Preference By Month
Figure 8. Chisana caribou ecotype preference by month as determined from GPS collar data.
Fecal Pellet Composition
Pellet composition by plant.
| Plant | Mean % composition | n |
|---|---|---|
| Classic Moss | 27.4 | 34 |
| Cladonia/Cladina lichens | 26.7 | 34 |
| Cetraria/Dactylina lichens | 7.6 | 34 |
| Alectoria/Bryoria lichen | 7.5 | 34 |
| Foliose lichen | 5.0 | 34 |
| Carex spp. | 4.1 | 34 |
| Equisetum spp. | 4.1 | 34 |
| Aulacomnium Moss | 2.8 | 33 |
| Dryas spp. leaf | 2.3 | 26 |
| Poa arctica | 1.7 | 30 |
| Empetrum nigrum leaf | 1.5 | 27 |
| Betula spp. stem | 1.3 | 19 |
| Sphagnum sp. Moss | 1.3 | 30 |
| Polytrichum juniperinum | 1.2 | 22 |
| Salix spp. leaf | 1.1 | 21 |
Total records = 15.
Pellet percent composition by plant form.
| Form | Mean % composition | n |
|---|---|---|
| Moss | 9.0 | 121 |
| Lichen | 8.8 | 185 |
| Sedge/Rush | 2.5 | 61 |
| Forb | 1.5 | 123 |
| Grass | 1.0 | 118 |
| Shrub | 0.9 | 254 |
Total records = 6.
Discussion
Vegetation Cover
Lichen Layer Thickness and Biomass
Litter Depth and Composition
Caribou Ecotype Preference
Fecal Pellet Composition
Appendix A: Sampling Plots
List Of Plots
| Plot | Ecotype | EstablishDate | Elevation (m) | Latitude | Longitude |
|---|---|---|---|---|---|
| ADD01 | Alpine Dwarf Shrub/Barrens | 2016-07-17 | 1420.039 | 62.08734 | -141.8660 |
| ADD04 | Alpine Dwarf Shrub/Barrens | 2016-07-16 | 1755.171 | 61.93552 | -141.9120 |
| ADD05 | Alpine Dwarf Shrub/Barrens | 2016-07-17 | 1778.203 | 62.02330 | -141.4990 |
| ADD06 | Alpine Dwarf Shrub/Barrens | 2016-07-15 | 1634.397 | 61.93408 | -141.1590 |
| ADD07 | Alpine Dwarf Shrub/Barrens | 2015-07-27 | 1430.455 | 61.96506 | -141.7315 |
| ADD09 | Alpine Dwarf Shrub/Barrens | 2015-07-11 | 1553.244 | 62.25612 | -142.4690 |
| ADD10 | Alpine Dwarf Shrub/Barrens | 2015-07-27 | 1571.130 | 61.97373 | -141.5712 |
| ADD12 | Alpine Dwarf Shrub/Barrens | 2016-07-16 | 1791.005 | 61.89077 | -141.8270 |
| ADD13 | Alpine Dwarf Shrub/Barrens | 2015-07-27 | 1930.868 | 62.06235 | -142.1865 |
| ADD14 | Alpine Dwarf Shrub/Barrens | 2015-07-27 | 1285.111 | 61.97960 | -141.4711 |
| ADD15 | Alpine Dwarf Shrub/Barrens | 2016-07-15 | 1971.142 | 61.93471 | -141.5820 |
| ASD01 | Alpine Sedge-Dwarf Shrub | 2015-07-27 | 1352.073 | 61.89981 | -141.1643 |
| ASD02 | Alpine Sedge-Dwarf Shrub | 2015-07-12 | 1327.991 | 61.75152 | -141.3301 |
| ASD03 | Alpine Sedge-Dwarf Shrub | 2016-07-15 | 1312.164 | 61.75139 | -141.0540 |
| ASD05 | Alpine Sedge-Dwarf Shrub | 2015-07-26 | 1356.013 | 61.95466 | -142.0188 |
| ASD07 | Alpine Sedge-Dwarf Shrub | 2015-07-21 | 1314.135 | 61.81529 | -141.6402 |
| ATS01 | Alpine Meadow | 2015-07-13 | 1634.906 | 61.89196 | -141.5397 |
| ATS02 | Alpine Meadow | 2015-07-28 | 1195.729 | 62.00252 | -141.2674 |
| ATS04 | Alpine Meadow | 2015-07-12 | 1620.190 | 61.84752 | -141.4842 |
| ATS06 | Alpine Meadow | 2015-07-29 | 1477.492 | 62.05676 | -141.7929 |
| RS03 | Riverine | 2015-07-11 | 1194.302 | 62.13432 | -142.3045 |
| RS04 | Riverine | 2015-07-29 | 1332.846 | 62.09194 | -141.8064 |
| RS05 | Riverine | 2015-07-11 | 1073.724 | 62.14444 | -142.1755 |
| SWB01 | Subalpine Wood/Shrub | 2015-07-28 | 1316.677 | 62.00130 | -141.1361 |
| SWB06 | Subalpine Wood/Shrub | 2016-07-16 | 1560.109 | 62.09998 | -142.0030 |
| SWB07 | Subalpine Wood/Shrub | 2015-07-21 | 1478.406 | 61.89189 | -141.4501 |
| SWB08 | Subalpine Wood/Shrub | 2016-07-17 | 1307.897 | 62.01965 | -141.3341 |
| SWB09 | Subalpine Wood/Shrub | 2015-07-11 | 1182.380 | 62.18664 | -142.3202 |
| SWB10 | Subalpine Wood/Shrub | 2015-07-12 | 1209.707 | 61.79014 | -141.4875 |
| SWB12 | Subalpine Wood/Shrub | 2015-07-26 | 1403.939 | 61.96335 | -141.9544 |
| SWB14 | Subalpine Wood/Shrub | 2015-07-13 | 1232.468 | 62.15854 | -142.2435 |
| SWB15 | Subalpine Wood/Shrub | 2016-07-16 | 1508.760 | 62.00719 | -141.7080 |
| WSP01 | Forest | 2015-07-28 | 1011.836 | 61.92991 | -141.0560 |
Elevational Profiles
Table 12. Sampling plots elevation profile by ecotype. Error bars represent the minimum and maximum plot elevation for each ecotype.
Figure 8. Sampling plots elevation profile by plot.